Creep laws for the mantle of the Earth

The analyses of glacial rebound data by Gathles and by Peltier and Andrews have led them to the conclusion that the flow law of the mantle of the Earth is Newtonian and that the viscosity is essentially a constant (1022 P) throughout the mantle. In this paper it is concluded that no large strain, steady-state creep process in mantle rock can account for a Newtonian, constant viscosity mantle. It is suggested that small strain, transient creep and not steady-state creep is involved in the isostatic rebound phenomenon. Since convective motion in the mantle involves large creep strains, conclusions about the effective viscosity of mantle rock undergoing such flow that is based on isostatic rebound data are likely to be wrong. If Post’s and Carter & Mercier’s laboratory results of the stress dependence of the grain size in a mantle type rock are representative of the actual grain sizes in the mantle, power law creep is almost certainly the creep law that governs convective flow in the mantle.